1. Field of the Invention
The invention generally relates to IP Multimedia Subsystem (IMS) networks and, more specifically, to IMS networks with dynamic call models.
2. Discussion of Related Art
Commonly deployed wireless communication networks, usually referred to as 2.5G networks, support both voice and data services. Typically, mobile handsets are connected to a Base Transceiver Station (BTS) using a Radio Access Network (RAN) that uses a modulation scheme such as CDMA (Code Division Multiple Access) or GSM (Global System for Mobile communications). The BTSs are connected via fixed links to one or more Base Station Controller (BSC) and the BSCs are aggregated into switches called Mobile Switching Centers (MSCs). The MSC is connected to the Public Land Mobile Network/Public Switched Telephone Network (PLMN/PSTN), typically through a gateway switch called the Gateway Mobile Switching Center (GMSC). Sometimes the term “core network” is used to collectively describe the MSC, GMSC and associated network elements. Voice traffic uses the so called circuit switched paradigm of communications in which circuits are assigned, i.e., dedicated, to a call for its entire duration; the voice traffic is carried using Time Division Multiplexing (TDM) switching technology. Signaling traffic uses Signaling System 7 (SS7) typically as out of band circuits.
With the advent of Internet Protocol (IP) networking, IP data service is offered to wireless clients by an overlay data network in which a packet control function (PCF) is introduced at the BSC level to connect BSCs to an IP-routed network. The PCF is responsible for packetization of RAN traffic. On the inbound side (core network to RAN) the PCF takes IP packets and reorganizes them for transmission as frames over the radio transport protocol. On the outbound side (RAN to core network) the PCF packetizes radio protocol frames to IP packets. Data connections are handled by this overlay network and the MSC is used primarily to handle circuit switched voice calls.
The development of Voice over IP (VoIP) technology has resulted in the MSC being re-designed to handle packet switched voice traffic along with existing circuit switched traffic. This new architecture is called a soft switch network. The legacy switch is disaggregated into a control and multiplicity of media gateway (MGW) components. The control component (sometimes called the soft switch) uses an open control protocol called the Media Gateway Control Protocol (MGCP) to manage the MGW. The MGW itself has the ability to accept both packet and circuit switched traffic and convert one to the other, under the control of the soft switch. It is thus possible in 2.5G networks to carry both circuit switched and packet switched traffic.
It is widely believed that wireless communications will soon be dominated by multimedia services. This has resulted in new RAN technologies and the resulting networks are called 3G networks. The transition of 2.5G to 3G networks emphasizes packet traffic and new architectures have been proposed to handle multimedia sessions, such as Quality of Service (QoS).
A defining characteristic of 2.5G/3G multimedia services is that since the handset can send or receive IP data packets at any time, the IP context of the handset is maintained as long as the handset is powered on and connected to the network. This is in contrast to traditional telephony where the state of a connection is maintained only while a telephone call is in progress.
In particular, in 3G networks the services are to be provided by so-called Application Servers. Consequently the connection between the service logic and the application server is a “stateful” connection that needs to be maintained for the duration of the service being used. Hence a very large number of stateful connections need to be maintained between the application server complex, hosted in the application domain, and the service logic complex hosted in the service logic domain, in a network servicing a large number of subscribers. Such stateful connections that cross administrative domains have high networking costs and are difficult to maintain operationally.
Typical of proposals for 3G network architecture is the IP Multimedia Subsystem (IMS) architecture, shown in FIG. 1, that uses Session Initiation Protocol (SIP) for control and signaling messages. SIP is an IP-based signaling protocol designed for multimedia communications. The IMS architecture introduces several control functions, i.e., functional entities, to manage the network. The legacy circuit switched traffic is handled by an Inter-working Function called the BGCF (Breakout gateway control function). The MGW is controlled by a new function called the Media Gateway Control Function (MGCF), and the media processing functions are performed by the Media Resource Function Processor (MRFP), which is controlled by the Media Resource Control Function (MRFC).
The basic call server called the Call State Control Function (CSCF) is logically partitioned into three functional entities, the Proxy, Interrogating and Serving CSCF.
The Proxy Call State Control Function (P-CSCF) is the first contact point for the handset, also referred to herein as the User Entity (UE) within IMS and provides the following functions:                1. Forward SIP register request received from the UE        2. Forward SIP messages received from the UE to the SIP server        3. Forward the SIP request or response to the UE        4. Detect and handle an emergency session establishment request        5. Generate Call Detail Records (CDRs)        6. Maintain Security Association between itself and each UE        7. Perform SIP message compression/decompression        8. Authorize bearer resources and QoS management        
The Interrogating CSCF (I-CSCF) is mainly the contact point within an operator's network for all IMS connections destined to a subscriber of that network operator, or a roaming subscriber currently located within that network operator's service area. It provides the following functions:                1. Assign a S-CSCF to a user performing SIP registration        2. Route a SIP request received from another network towards the S-CSCF        3. Obtain from Home Subscriber Server (HSS) the Address of the S-CSCF        4. Forward the SIP request or response to the S-CSCF as determined above        5. Generate CDRs        
The Serving CSCF (S-CSCF) actually handles the session states in the network and provides the following functions:                1. Behave as SIP Registrar: accept registration requests and make its information available through the location server        2. Session control for the registered endpoints' sessions        3. Behave as a SIP Proxy Server: accept requests and service them internally or forward them on        4. Behave as a SIP User Agent: terminate and independently generate SIP transactions        5. Interact with application servers for the support of Services via the IMS Service Control (ISC) interface        6. Provide endpoints with service event related information        7. Forward SIP message to the correct CSCF        8. Forward the SIP request or response to a BGCF for call routing to the PSTN or CS Domain        9. Generate Call Detail Records.        
The P-CSCF is the first point of contact for a UE (handset) in an IMS network. The I-CSCF then helps in establishing which S-CSCF “owns” the UE.
FIG. 2 is a signaling diagram 200, showing the call flow for a UE when it first establishes contact with an IMS network. The UE sends a “register” request to the proxy. Assuming the proxy determines that the UE is registering from a visiting domain, it queries the DNS to find the I-CSCF in the UE's home domain. The proxy then sends the registration information to the I-CSCF. The HSS checks if the user is already registered and sends the address of the S-CSCF in response. An authentication process now ensues in which the UE is challenged to provide valid authentication vectors. Once the authentication procedure is completed, the S-CSCF informs the HSS that the UE is registered.
The HSS provides initial filter codes (IFCs) to the S-CSCF. The IFC, in effect, maps the service codes with various application servers (ASs). Thus, if the UE later issues a service request or if the service is otherwise triggered the mapped AS will be invoked. The IFC is effectively the “call model” for the UE. These call models are static objects downloaded during registration from the HSS. Every UE in the domain of the S-CSCF will, if they have the services enabled at all, have the same application servers (ASs) mapped to the same services. For example, push-to-talk service for each and every UE having such service will point to the same AS or point to an AS with identical service logic to provide the identical push-to-talk functionality.
Registered UEs may use services by initiating a new session establishment procedure depicted in FIG. 3. The figure shows a session establishment request originating with a S-CSCF (called O-SCSCF) or I-CSCF (called O-ICSCF). This request is routed to the “terminating” S-CSCF (T-SCSCF), which consults the callee's service profile. Based on the service profile of the originating registered user, the T-SCSCF sends an IMS service control request (ISC) to the corresponding application server (T-AS) that can handle this service request. The T-AS provides the service to the callee and terminates the session and the S-CSCF terminates the application activation process.
As an illustrative example, consider the case of voice mail in which callers to a certain user may leave a voice message if the called user does not respond to the call. This voice mail service is provided by an application server (AS) dedicated to this service and having service logic to provide such functionality. The S-CSCF transfers control to the voice mail application server when a certain service point trigger (SPT) occurs, i.e., an event occurs that causes a trigger within the SPT to “fire.” The IFCs that provide trigger points to the service logic of the S-CSCF are downloaded into the S-CSCF during user registration at session initiation time and remain fixed for the duration of the session. The service profile described above that is consulted by the T-SCSCF is a static object in the sense that the information contained in it is defined once at the time of service inception.
Since government regulations restrict use of the electromagnetic spectrum to license holders, such license holders have traditionally been the only organizations to offer wireless services. However, the industry has recently seen the advent of Mobile Virtual Network Operators (MVNOs) that use the license holder's underlying network facilities to offer services. In essence, this is a business agreement that allows the MVNO to use the network operator's facilities without owning the network infrastructure or spectrum license. Thus, wireless services can be offered by several MVNOs (as virtual service providers) using the same underlying network infrastructure. One problem with such an arrangement is that service differentiation is not possible amongst the various MVNOs. Thus, if two MVNOs, MVNO1 and MVNO2, both provide voice service, that service will be indistinguishable to end users of the voice service.
The service that is offered on a 3G network is determined by the service logic that is executed by the network infrastructure. In prior art systems, the service logic is executed by the network infrastructure, i.e., the network facilities and equipment. Therefore, all users of the network are tied to the same service logic, namely that of the network infrastructure, and consequently are constrained to offer the same service experience. In the example referred to above, MVNO1 and MVNO2 are both tied to the service logic provided by the network infrastructure, and they are not able to differentiate their services. Even though the customer base can be distinguished by the network infrastructure, for example for billing purposes, the service characteristics for all customers cannot be distinguished unless different service logic is executed for different sets of users.